Task 1: Effect of Coal Quality
The objective of this task is to assess if fuel selection is an important factor determining ash quality. Work on this task will involve each of the three participating organizations. Ash samples from three coals will be generated under identical firing conditions in the pilot furnace at the University of Utah, and the matching ash and coal samples sent to Brown. Additional matching sets of coal and ash will be obtained from commercial-scale firing at Southern Company. The ashes will be characterized for LOI and surfactant adsorption activity under standard conditions and trends with fuel type identified. At the same time, chars will be prepared from the matching coal set under standard conditions in a laboratory furnace and also characterized for surfactant adsorptivity. A variety of standard conditions may need to be explored. The combined data set will be analyzed to determine cross correlations between ash behavior, standard laboratory char behavior, and parent coal properties. Our goal is to be able to anticipate ash behavior either (a) from coal properties directly, or (b) from the properties of chars made by a simple laboratory procedure. Either could be the basis for a coal quality index -- one based on fuel properties and the other based on a simple screening test.

To build on the results above, the same simple laboratory procedure will also be used to make chars from a large set of solid fuels (currently under study at Brown). This set of solid fuels includes US. coals of various rank, international coals, and alternate fuels from wood, agricultural residues, petroleum cokes, and organic model compounds. The surfactant adsorptivity of these chars will be determined under standard conditions and correlations developed with parent fuel properties. Aside from one published measurement on petroleum coke cofiring ashes this will be the first data taken on the surfactant adsorptive properties of alternative fuels. Additional char characterization will be done as needed, including measurement of surface areas, pore size distributions, and surface polarity.

The goal of this work will be to develop correlations or simple tests that can be used to predict ash behavior for the wide variety of fuels of interest in today's utility sector. It is likely that strong correlations will be seen with basic fuel properties (in the same way that oxidation reactivity is showing strong correlations with parent fuel properties in a parallel study underway at Brown).

Task 2: Effect of Combustion Conditions and Storage History
The goal of this task is to understand how combustion conditions, especially low-NOx firing configurations affect the key unburned carbon properties and to characterize for the first time changes in activity occurring during ash storage and handling. This task will also generate as a byproduct large samples of primary PM2.5 from well-defined pilot-scale combustion conditions.

The pilot-scale furnace at the University of Utah will be operated under both high swirl high-NOx and two staged combustion low NOx conditions to generate fly ash samples for analysis. The coals will be a Powder River Basin sub-bituminous, an eastern bituminous steam coal, and a Utah bituminous coal. The combustion conditions will be selected to generate low carbon and high carbon ash samples from the same coal with the same furnace-firing rate and exit stoichiometry.

Bulk mixed fly ash will be collected using a bag filter that collects ash from the entire furnace exhaust producing sample at a rate of 100-200 g/hr. Size fractionated flyash, d>10 m, 10-2.5 m will be produced using a cyclone, 20-jet preseparator, and final filter. Since most of the mass is in the large particles, less than 5g of PM2.5 ash may be produced from a day of furnace operation.

The ash will be sampled with an Andersen cascade impactor to obtain data on carbon content as a function of particle size and to obtain size-fractionated samples for electron microscopy. The mass of the cascade impactor samples will be too small for use in concrete additive assays but will provide useful data on the variation of ash composition and morphology by particle size.

Particle characterization will consist of measuring surface area by nitrogen adsorption, measuring the fraction of carbon in the form of soot and of char in the bulk ash samples, and examining morphology of the mineral and carbonaceous particles by electron microscopy. Kinetic modeling studies will apply existing ash transformation and carbon burnout models to the specific time-temperature-oxygen conditions of the pilot furnace during the sample generation. The goal will be to use these models to gain insights into the relationship between combustion conditions and the ash characteristics that are important for fly ash recycling.

As stated earlier, it is likely that air exposure of fresh ash samples will reduce their surfactant adsorptivity by mild oxidation which introducing polar functionalities to the carbon surfaces. This effect likely occurs naturally in ash handling and storage systems, but in an uncontrolled and unverified manner. We wish to determine if this effect is real and can be exploited in some way for improvement of ash properties. Samples from the furnace will therefore by initially stored under a N2 blanket and subsequently exposed to atmospheric conditions for specified time periods prior to measurement of the surfactant activity. If significant effects are observed over reasonable times, this will provide a very economical strategy for up grading marginal ash streams in industry.

Task 3: Beneficiation with Ozone
The goal of this task is to carry out critical laboratory experiments in parallel to, and in support of, the development of ozonation as a commercial ash beneficiation process. A carbon-rich fraction of a commercial fly ash will be obtained by size classification and used as the base material for this study. The reduction in inorganic content is expected to have no important effect on the O₂/C reaction, but will reduce interferences in the characterization of the carbon surfaces. The carbon-rich ash will be ozonated as a function of time and ozone concentration to yield a set of samples with varying degrees of surface polarity. The standard surfactant adsorptivity will be measured as an indicator of the extent of effective treatment for concrete use. A battery of techniques to determine the number and nature of the oxygen complexes present will then be used to characterize the samples. These techniques include:

• FTIR diffuse reflectance (using a recently acquired instrument at Brown)
• Thermal programmed desorption (TPD) using either FTIR or mass spec analysis of the desorption complexes.
• Surface polarity by micro-flow-calorimetry

The use of XPS will be explored in a collaboration arrangement with one or more outside laboratories. The thermal desorption tests will also provide information on the temperature range for stability of the complexes, which may have implications for the proper storage and treatment of ozonated ash.

Early tests showed air oxidation to also have some beneficial effect on surfactant adsorptivity, but not to the same degree as ozonation. Some additional characterization will therefore be done with air-oxidized carbon surfaces to understand the origin of the different behavior of O₃ (at near ambient temperature) and O₂ (at elevated temperature). Side reactions between ozone and the inorganic portion of the ash may also be important, and experiments will be carried out with pure inorganic compounds to understand these alternate reaction pathways. The critical effect of temperature will be investigated, as will the effects of carbon particle size and origin (parent fuel).